Protocol Characteristics

To enable communication over heterogeneous systems, a common standard such as the OSI reference model is used. This model contains seven layers, and each layer performs a specific set of functions. Each layer communicates with the corresponding layer of the other system. These protocols are components of layers of a protocol suite. A single layer may include one or more protocols. For example, Layer 2 of the OSI model includes more than one protocol—PPP and ATM. Table 2.3 lists the seven layers of the OSI Model.

Table 2.3: The Seven Layers of the OSI Model

Layer	Function
Application	Is concerned with the user interface and the application program interface, such as FTP or Telnet. It is also concerned with the management of the data transfer process.
Presentation	Is concerned with the data format received from the Application layer, and the format required by the destination node.
Session	Establishes a logical session between the source and destination nodes of the network. In addition, this layer maintains and synchronizes the established session until data is transferred to the destination node.
Transport	Provides end-to-end connectivity between the source and the destination nodes using acknowledgments and flow control mechanisms.
Network	Provides logical and physical addresses to data packets being transferred and routes them along the best route available.
Data-link	Facilitates data transfer through the communication channel by consolidating data into frames. This layer also checks for transmission errors and applies mechanisms for correcting it.
Physical	Provides the actual mechanical and electrical interface for data transmission. Data transmission at this layer is in bits and bytes.

The most commonly used protocol suite of the Transport and Network layers is TCP/IP. Using the packet-switching technology, the TCP/IP protocol suite enables data transfer between dissimilar networks. This was developed by the Defense Advanced Research Projects Agency (DARPA) and Stanford University. Together, these two institutions developed the Internet Protocol (IP) suite, which consists of the TCP and IP protocols.

The TCP/IP protocol suite provides communication among complex networks. For example, the network of an organization ABC is connected through a WAN link with the network of its partner, XYZ. ABC uses the TCP/IP protocol suite, whereas XYZ is using NetWare protocols. To enable data transfer and communication between these two networks, TCP/IP provides data translation and compatibility features.

Table 2.4 displays protocols of the TCP/IP protocol suite that are applicable to various layers of the OSI layers.

Table 2.4: Protocols for OSI Layers

OSI Layer	Protocols
Network layer	IP and ICMP
Transport layer	TCP and UDP
Session, Presentation, and Application layers	RIP, IGRP, and OSPF

Note

Other protocols, such as FTP, TELNET, SMTP, and SNMP are also applicable to the Session, Presentation, and Application layers.

Protocols of the Data-link Layer

The Data-link layer arranges raw data transmitted by the Physical layer into data frames, transmits them sequentially, and processes acknowledgment frames. The protocols used by the Data-link layer are:

• Point-to-Point Protocol (PPP)

• Synchronous Data Link Control (SDLC)

• High-Level Data Link Control (HDLC)

Point-to-Point Protocol

PPP manages error detection, permits authentication, supports multiple protocols, negotiates IP addresses at the time of connection, and solves the problem of assigning and managing IP addresses. PPP consists of:

Framing Flag: Defines the beginning and end of each frame.

Link Control Protocol (LCP): Establishes, configures, and tests the data-link connection.

Network Control Protocols (NCP): Establishes and configures different Network layer protocols by having a different NCP for every Network layer.

To establish communication over a PPP link:

1. The Network layer receives a series of NCP data packets when the link is established.

2. Packets from the Network layer are sent over the data link after the Network layer protocols are configured.

Figure 2.2 depicts a PPP frame format.

Figure 2.2: The PPP frame format.

Table 2.5 lists functions of each field of the PPP format.

Table 2.5: Fields of PPP Frame

Field	Description
Flag	Indicates the beginning or end of each frame. All the PPP frames begin with the binary value 01111110.
Address	Contains the binary value 11111111, which is a common broadcast address. Because PPP does not allocate a separate node address, it supports only one connection between two nodes.
Control	Contains the binary value 00000011, which indicates an un-numbered frame, because PPP does not offer reliable transmission using sequence numbers and acknowledgments.
Protocol	Identifies the protocol in the Data field. For example, a 0 bit indicates that the protocol is a Network layer protocol, such as IP and IPX.
Data	Contains the data for transfer. This field is of variable length, which is decided using LCP during the communication line setup. However, the default length of this field is 1500 bytes.
Checksum	Checks for any errors in data transfer. This field is normally 2 bytes. However, 4 bytes can be used for more efficient error

Figure 2.3 illustrates the different phases of establishing and terminating a PPP link.

Figure 2.3: The different phases of a point-to-point link connection.

Table 2.6 lists the different phases of a PPP link connection.

Table 2.6 : Phases of the PPP Link

Phase	Description
Link Dead	The communication line is dead; that is, no Physical layer carrier is available. In addition, no Physical layer connection exists.
Link Established	The Physical layer connection is established, that is, negotiation of an LCP option begins. If it is successful, the link moves to the Link Authenticated phase.
Link Authenticated	The two communicating nodes check the identity of the other.
Link Networked and Opened	The suitable NCP protocol is invoked to configure the Network layer. If successful, data transfer takes place.
Link Terminated	In this phase, the communication link is terminated, after data transfer is complete.

Synchronous Data Link Control (SDLC)

SDLC is a synchronous bit-oriented operation protocol, which can be used with links and topologies such as:

• Point-to-point link

• Half-duplex transmission

• Full-duplex transmission

• Circuit-switched networks

• Packet-switched networks

SDLC has two types of network nodes: primary and secondary. The primary node manages the operation of one or more nodes, also known as secondary nodes. It also establishes and terminates links with the secondary node. The primary node requests that the secondary nodes send the data in a predetermined sequence. The primary nodes manage the secondary nodes. The secondary nodes can send data to a primary node only after getting permission from the primary node, as depicted in Figure 2.4.

Figure 2.4: Functions of primary and secondary nodes.

Table 2.7 shows the four main configurations in which primary and secondary nodes can be configured.

Table 2.7 : Configuration of Primary and Secondary Nodes

Configuration	Description
Point-to-point	Connects two nodes, primary and secondary.
Multipoint	Connects one primary node with multiple secondary
Hub Go-ahead	Involves two channels, incoming and outgoing. The primary node uses the outgoing channel to communicate with the secondary nodes. The secondary node uses the inbound channel to send data to the primary node.
Loop	Connects a primary node to the first and last secondary nodes. The intermediate nodes pass messages to one another for communicating with the primary node.

Figure 2.5 displays the SDLC frame format.

Figure 2.5: The frame format of the SDLC protocol.

Table 2.8 states the description of each field in the SDLC frame format.

Table 2.8 : Fields of the SDLC Frame

Field	Description
Flag	Indicates the beginning or end of each SDLC frame.
Address	Contains the address of the secondary node involved in current communication.
Control	Uses three different formats: Information, Supervisory, and Unnumbered. These formats depend on type of SDLC frame used.
Data	Contains the data to be communicated and is of variable length.
CRC	Performs error detection during data transfer.

High-Level Data Link Control (HDLC)

HDLC is a multipurpose data link control protocol that can be used on both point-to-point and multipoint links. Like SDLC, it supports full-duplex transparent-mode transmission.

HDLC is similar to SDLC in that their frame formats are identical. In addition, both protocols support a full-duplex operation. Table 2.9 lists the differences between HDLC and SDLC.

Table 2.9 : Differences Between SDLC and HDLC

SDLC	HDLC
Supports hub go-ahead and loop configurations.	Does not support hub go-ahead and loop configurations.
Does not have the 4-byte checksum option or CRC.	Contains 4-byte checksum option.
Supports only one transfer mode—the	Supports three transfer modes: NRM, Asynchronous Response Mode (ARM), and
Normal Response Mode (NRM). In this mode, secondary nodes cannot communicate with the primary node until the primary node permits them.	Asynchronous Balanced Mode (ABM). In ARM, secondary nodes can start communicating with the primary node without obtaining its permission. ABM introduces a combined node that can serve as a primary or secondary node with respect to the situation.

Figure 2.6: The HDLC Frame Structure

Protocols of the Network Layer

IP is the primary protocol used on the Network layer. This protocol routes data packets from the source to the destination node in a network. Therefore, this protocol is also called a routing protocol. The IP protocol fragments data packets into datagrams as the data packets are transmitted to the destination node. Each datagram is 1500 bytes and is fragmented into smaller segments as it is transmitted. At the destination node, IP reassembles the fragmented datagrams to form the original data packet. In addition, IP identifies and detects errors that may occur during data transmission.

When data packets are fragmented into datagrams, each datagram has an associated header attached to it. This header contains vital information about the datagram, such as the source address, destination address, and protocol version. The header of a data packet is of 32 bits, which includes 20 bytes of fixed length fields. The remaining fields are of variable length. Figure 2.7 displays the header datagram of IP.

Figure 2.7: The IP header datagram.

The fields of the IP header datagram are:

Version: Indicates the current version of the IP protocol. If the current version of IP is 4, it is called IPv4. The size of this field is 4 bits.

IHL: Indicates the length of the IP header. By default, the length of the IP header is 32 bits. The size of this field is 4 bits.

Type of Service: Indicates the manner in which the upper-layer protocol wants the datagram to be handled. Using this field, the priority level for a particular datagram can be specified. The size of this field is 8 bits.

Total Length: Specifies the total length of the data packet. The total length includes the length of the datagram header. The size of this field is 16 bits.

Identification: Specifies the identification number of the current datagram. This identification number enables reassembling of datagrams at the destination node. The size of this field is 16 bits.

Flags: Indicates the fragmentation and the total bits in the entire data packet. The size of this field is 3 bits. The two lower bits of the three bits control the fragmentation of the datagram. The first of these two bits indicates the possibility of the current packet being fragmented. If this bit is set to 1, it indicates that the data packet cannot be fragmented; if the bit is 0, the data packet can be fragmented. The second bit of the two lower order bits specifies whether the current packet is the last in a series of packets. If this bit is 0, it indicates that the current packet is the last of the series; if the bit is 1, there are more packets following the current packet. The third bit of the three bits is a high-order bit and is set to 0.

Fragment Offset: Indicates the exact location of the current datagram in the entire data packet. The size of this field is 13 bits.

Time-to-Live (TTL): Specifies the total time for which a datagram is active and valid. A counter is associated with this time, which gradually decrements to 0, at which point, the datagram is discarded and considered invalid. The size of this field is 8 bits.

Protocol: Specifies an integer value, which indicates that an upper-layer protocol has received the incoming data packets after the IP protocol processing is complete. For example, if the value in this field is 6, TCP will receive the processed incoming data packets.

Header Checksum: Detects errors in the header of the data packet. The size of this field is 16 bits.

Source Address: Specifies the address of the source host. The size of this field is 32 bits.

Destination Address: Specifies the address of the destination host. The size of this field is 32 bits.

Options: Provides additional information, such as security information or information related to a path adopted by the data packet. This field can specify the complete path that a data packet needs to follow. The size of this field is 32 bits.

Data: Consists of information related to the upper layers.

In addition to the header information, addressing schemes used at the Network layer are an important part of routing IP data packets.

Addressing at the Network Layer

An IP address is a unique 32-bit Layer 3 address that uniquely defines a host in the network and allows it to participate in a TCP/IP network. The IP address is a 32-bit series divided into four groups of eight bits (octets) each. This series is written in a decimal notation with numbers ranging from 0 to 255.

Note

The IP address is assigned by the network administrator and differs from a MAC address, which is allocated by the hardware manufacturer.

An IP address is made up of 32 bits. This address is divided into four bytes or octets. Table 2.10 shows a 32-bit IP address in a structured addressing scheme.

Table 2.10 : The 32-bit IP Address

IP Address Notation	Address
Dotted decimal	192.168.12.1
Binary	11000000.10101000.00001100.00000001
Subnet mask in dotted decimal	255.255.255.0
Subnet mask in binary	11111111.11111111.11111111.00000000

Each network can again be subdivided into a number of subnets, depending on the requirement. Finally, each host within the same network or subnet should have a unique IP address, which identifies it in the entire network. This is also called the host address. In Table 2.10, the host address is 192.168.12.1. The network part of the IP address identifies the network to which the particular host belongs. In order to communicate, all hosts should have the same network address.

When a greater number of hosts and lesser number of networks are required, more bits are allocated to the hosts and less to the networks, and vice versa. For example, 192.168.2.4 is an IP address with a subnet mask of 255.255.0.0. The first two bytes or octets represent a network address, and the last two octets are the host address, as shown in Figure 2.8.

Figure 2.8: Allocation of IP addresses.

The number of bits assigned to the network and host portions depends upon the number of networks to be configured. In a public network, InterNIC assigns this network number. To identify the number of bits that determine the network portion, a subnet mask has to be used along with the IP address. The subnet mask determines the demarcation point between the network and the host portion of the IP address. This is explained later in the chapter.

Classes of IP Addressing

The requirement to allocate IP addresses to networks of varying sizes was addressed by introducing the concept of address classes. The hierarchical model of the IP addressing scheme gave rise to different classes of IP addressing. InterNIC assigns the classes of IP addresses to an internetwork with respect to the size of the network. This is to avoid any confusion during the allocation and distribution of IP addresses. There are five different address classes:

• Class A

• Class B

• Class C

• Class D

• Class E

Internetworks are divided into three sizes:

Big internetworks: Assigned a larger number of hosts and fewer networks. These organizations are allocated Class A address.

Medium internetworks: Assigned a requirement of hosts and subnetworks in between the big and small internetworks. These organizations are allocated Class B address.

Small internetworks: Assigned a smaller number of hosts and larger number of subnetworks. These organizations are allocated Class C address.

Table 2.11 lists the specifications and options associated with different classes.

Table 2.11 : Class Requirement and Available Options

Class	Purpose	Maximum Networks	Maximum Hosts
A	Large organizations	127	16,777,214
B	Medium-size organizations	16,384	65,543
C	Small organizations	2,097,152	254

The Internet community defined a set of rules in the hierarchical IP addressing scheme. For addresses in Class A, the leading bits of the first octet should always start with 0. The leading bits should be 10 for Class B, 110 for Class C, 1110 for Class D, and 1111 for Class E. Table 2.12 shows the address ranges of different classes.

Table 2.12 : Leading Bits and Address Ranges of Classes

Class	Leading Bit	Address - Range
A	0	1.0.0.0 -127.255.255.255
B	10	128.1.0.0 - 191.254.0.0
C	110	192.0.0.0 - 223.255.255.255
D	1110	224.0.0.0 - 239.255.255.255
E	1111	240.0.0.0 - 254.255.255.255

Class A

Class A addresses range from 1.0.0.0 to 126.0.0.0, where the first octet represents the network portion, and the last three octets represent the host. The Class A address format is used for large organizations with networks supporting a large number of end users. The maximum number of networks possible with Class A addressing is 127, and the maximum number of hosts per network number is 16,777,214. The highest order of the network bits is always the most significant and defines the class of the network. In the case of Class A networks, the highest order bit—the first bit of the first octet—is zero. Figure 2.9 depicts the Class A addressing format.

Figure 2.9: The addressing format of Class A.

Class B

Class B addresses range from 128.1.0.0 to 191.254.0.0, where the first two octets represent the network portion, and the other two octets represent the host. The Class B address format is used for networks of mid-sized organizations. The maximum number of networks possible with Class B addressing is 16,384, and the maximum number of hosts per network is 65,543. The highest order of the network bits is always 10. The first bit of the first octet is set to 1, and the second bit is set to 0. Figure 2.10 depicts the Class B addressing format.

Figure 2.10: The addressing format of Class B.

Class C

Class C addresses range from 192.0.1.0 to 223.255.254.0, where the first three octets represent the network portion, and the last octet represents the host. The Class C addressing format is used for small organizations with networks supporting a large number of users. The maximum number of networks possible with Class C addressing is 2,097,152, and the maximum number of hosts per network number is 254. The highest order of the network bits is always 110. Figure 2.11 depicts the Class C addressing format.

Figure 2.11: The addressing format of Class C.

Classes D and E

Unlike classes A, B, and C, classes D and E are not for commercial use. Class D addresses are used for multicast groups and range from 224.0.0.0 to 239.255.255.255. Class E addresses are used for experimental purposes and range from 240.0.0.0 and 254.255.255.255.

Table 2.13 shows a breakup of IP addresses of Class A, B, and C in binary format.

Table 2.13 : IP Addresses in Binary Format

Network Number (Dotted decimal)	Network Number (Binary)
10.1.1.0	00001010.00000001.00000001.00000000 (Class A)
150.5.5.0	10010110.00000101.00000101.00000000 (Class B)
192.1.1.0	11000000.00000001.00000001.00000000 (Class C)

In Table 2.13, note that Class A address has 0 in the first bit of the first octet, Class B has 10 in the first two bits of the first octet, and Class C has 110 as the first three bits of the first octet. Table 2.14 lists the characteristics of all the five classes of IP addresses.

Table 2.14 : Characteristics of Classes A, B, C, D, and E

Class	Format	Purpose	Leading Bit	Address Range	Maximum Networks	Maximum Hots
A	N.H.H.H	Large organizations	0	1.0.0.0 – 126.0.0.0	127	16,777,214
B	N.N.H.H	Medium-size organizations	10	128.1.0.0 – 191.254.0.0	16,384	65,543
C	N.N.N.H	Small organizations	110	192.0.1.0 – 223.255.254.0	2,097,152	254
D	N/A	Multicast groups	1110	224.0.0.0 – 239.255.255.255	N/A	N/A
E	N/A	Experimental	1111	240.0.0.0 – 254.255.255.255	N/A	N/A

It is possible to connect networks with different classes of IP addresses. Figure 2.12 depicts a scenario in which the networks 10.1.1.0 (Class A), 150.5.5.0 (Class B), and 192.1.1.0 (Class C) are internetworked.

Figure 2.12: Internetworking of different classes of an IP network.

In Figure 2.12, the networks belong to three different classes of address: 10.1.1.0/8 (Class A), 150.5.5.0/16 (Class B), and 192.1.1.0/26 (Class C).

Routing at the Network Layer

The main function of the Network layer is to route data packets from the source to the destination nodes. IP performs this function by using the IP address of data packets. An IP address enables IP to identify the source and destination nodes. These source and destination nodes can be on different networks connected through routers.

For routing data packets on the appropriate path, the Network layer uses dynamic routing protocols, which contain routing software that decides the routes to be taken by each incoming data packet at specified intervals.

The Network layer uses the Internet Control Message Protocol (ICMP) and ICMP Router Discovery Protocol (IRDP) to route data packets on the appropriate path. In addition, protocols use the IP routing table, which specifies the destination address and the next-hop pair for the data packet. The next-hop pair specifies the address of an intermediate node, which is the next stop of the data packet from its current location to reach the specified destination address. The IP routing table does not specify the exact path of the data packet at the source node. It contains the address of the next intermediate node. The address of the next intermediate node is calculated by matching the destination address of the data packet with the address in the routing table of the current node. This type of routing process is known as dynamic routing.

Table 2.15 displays a sample routing table.

Table 2.15: Sample Routing Table

Destination Address	Next-Hop
36.10.0.0	45.32.22.1
79.3.0.0	35.32.10.9
17.12.3.0	45.32.22.2

From Table 2.15, you can infer that to reach subnet 10 of network 36, the next stop of the data packet is 45.32.22.1.

However, the IP routing table does not provide any mechanism to track whether data packets have reached the destination without any error, such as corruption of the data packet. To track errors related to data packets or routing failures, the Network layer uses ICMP.

ICMP

ICMP is a protocol used to track errors that corrupt data packets during transmission and routing failures that prevent data packets from reaching the destination node. In addition, ICMP reports transit errors to the source node.

To check for the destination node in a network, the ICMP protocol uses two commands:

ping: Performs the echo function that sends the data packet on a roundtrip between two hosts in a network. It provides details about packet forwarding and packet loss at each router and link in the path.

trace: Traces the path by sending router advertisement messages to identify the addresses of routers that are directly attached to the subnetworks, produces the command-line report output about each router that is crossed, and generates the Roundtrip Time (RTT) for each hop.

In addition, ICMP can also be used as a diagnostic tool to determine routing faults, security attacks, and misconfigurations within the network.

IRDP

IRDP uses router advertisements to identify addresses of routers that belong to directly attached subnets. Using IRDP, each router periodically broadcasts router advertisements to all hosts in a network. The hosts can use router advertisements to identify nearby routers and send requests to them.

Protocols of the Transport Layer

The Transport layer of the OSI model uses TCP and UDP protocols for transferring data through the network. TCP is responsible for the connection-oriented data transfer, whereas UDP provides connectionless data transfer.

TCP

TCP provides full-duplex and acknowledged services to the upper layers. In addition, TCP applies the flow-control mechanism to prevent a destination node from being flooded with data packets from a high-speed source node. TCP sends the data packets in the form of continuous and unstructured byte streams, which are uniquely identified by sequence numbers. Figure 2.13 displays a TCP packet format.

Figure 2.13: The header format of TCP.

The fields of the header format are:

Source Port: Identifies the source processes in the upper layers that receive TCP services. The size of this field is 16 bits.

Destination Port: Identifies destination processes in the upper layers that receive TCP services. The size of this field is 16 bits.

Sequence Number: Contains a sequence number assigned to the first byte of the data being sent. This sequence number uniquely identifies the data message being transmitted. The size of this field is 32 bits.

Acknowledgment Number: Contains the sequence number of the next byte of data being transmitted. The size of this field is 32 bits.

Data Offset: Indicates the number of 32-bit words in a TCP header. The size of this field is variable length.

Reserved: Is reserved for future use. The size of this field is 6 bits.

Flag: Specifies control information, such as establishing, terminating, and resetting a session. The size of this field is 6 bits.

Window: Specifies the size of the buffer space available at the source node to receive incoming data packets. The size of this field is variable length.

Checksum: Identifies if a data packet was corrupted during transmission. The size of this field is 16 bits.

Urgent Pointer: Contains a pointer to the last byte of any urgent data that is being transmitted to the destination node. The size of this field is 16 bits.

Options: Contains additional information about the data being transmitted to the destination node. The size of this field is variable length.

Data: Contains information about the upper layers. The size of this field is variable.

UDP

UDP is a simpler protocol than TCP, as it contains only four fields in its header. UDP is used where the source and the destination nodes do not require reliable and connection-oriented service. The fields of the UDP packet are:

Source Port: Identifies the source processes in the upper layers that receive UDP services. The size of the field is 16 bits.

Destination Port: Identifies the destination processes in the upper layers that receive UDP services. The size of the field is 16 bits.

Length: Specifies the total length of the UDP datagram, which includes a header and the user data. The size of the field is 16 bits.

Checksum: Identifies if a data packet was corrupted during transmission. A value of 0 in this field indicates that the checksum has not been calculated and used. This field is optional; its size is 16 bits.

Protocols of the Session, Presentation, and Application Layers

The Session, Presentation, and Application layers of the OSI model are concerned with the applications accessed by end users. Therefore, protocols supported by these layers perform several application-oriented functions, such as network management, file transfer and management, e-mail, and distributed network services.

Some protocols used by the Session, Presentation, and Application layers to ensure effective delivery of data packets to the destination node are:

• Routing Information Protocol

• Interior Gateway Routing Protocol

• Enhanced Interior Gateway Routing Protocol

Routing Information Protocol

The Routing Information Protocol (RIP) is suitable for small, homogenous networks and is relatively easy to implement. It has certain limitations that render it unsuitable for large and complex networks:

• Does not support classless routing (leads to waste of address space)

• Cannot be used in networks that require CIDR and VLSM

• Is not suitable for networks that require more than 15 hops

• Convergence is slow and is prone to routing loops

• Uses broadcasts for routing updates causing an increase in network traffic

RIP uses hop-count for its metric. A hop-count of 16 signifies an unreachable network. RIP sends request and response messages during routing updates. RIP uses UDP port 520 for all messages. RIP sends out request messages, using broadcasts when the RIP process begins.

Neighboring routers send response messages containing their routing tables, after receiving updates. On receiving the response message from its neighbor, a router checks to see whether the update is new or not. If the update is found to be new, it is entered in the routing table along with the address of the advertising router. If the entry already exists for that network, the updates are ignored unless one is received with a lower hop-count.

Note

A hop count between 0 and 15 is considered valid. If the hop count is 16, it would indicate a networkthat is unreachable.

RIP uses the major classful network number for route summarization, because it does not carry subnet mask routing information. Subnet masks enable efficient use of the IP addressing scheme. The lack of subnet mask information carried by RIP means that a router should assume that the subnet mask it has been configured with is effective for all the subnets.

When multiple paths can be used to reach a destination network, the router makes a choice based on the reliability of the routing information source, known as its administrative distance. The higher the value of administrative distance, the lower its reliability. RIP has an administrative distance of 120.

Note

RIP is defined in Request For Comment (RFC) 1058 and 1723.

The fields of the RIP header format are:

Command: Contains an integer with a value of 1 or 2. Value 1 indicates a request command. This command requests the responding network system to send all or part of its routing table to the destination node. Value 2 indicates a response command that responds to the request command and includes all or part of the routing table of the responding network system.

Version: Specifies the version of RIP, which can be either 1 or 2.

Address family identifier: Identifies the type of address family being used for the current RIP implementation. The size of this field is 16 bits.

Metric: Contains hop counts that a data packet needs to traverse before reaching the destination node. However, the maximum hop limit is 15. A destination that requires more than 15 hops is called unreachable.

In addition, RIP uses certain timers that regulate its performance. For example, the RIP routing update timer is set at 30 seconds, which ensures that every router sends its entire routing table to its neighbors.

Interior Gateway Routing Protocol

Interior Gateway Routing Protocol (IGRP) is a distance vector protocol, which is suitable for medium and large-sized networks. IGRP advertises the entire or a part of the routing table to neighbors via broadcasts. The metrics used in IGRP are a combination of delay, bandwidth, reliability, load, and Maximum Transmission Unit (MTU). IGRP has features such as split horizon, split horizon with poison reverse, route poisoning, hold-down timers, triggered updates, and count to infinity.

Being a distance vector protocol, IGRP advertises the entire or a part of the routing table to its neighbors every 90 seconds via broadcasts. If a route remains unreachable even after three consecutive updates, it is declared inaccessible. However, the inaccessible route is not removed from the routing table. It is only deleted after it fails to respond to seven consecutive routing updates.

The routes advertised by IGRP can be classified into:

Interior routes: For subnets connected to the interface of the router.

System routes: For networks from routers within the same AS. These networks do not include subnetworks.

Exterior routes: From networks outside ASs. These routes are used while defining a default gateway using the Gateway of Last Resort.

The message format of IGRP is depicted in Figure 2.14.

Figure 2.14: The message format of IGRP.

The descriptions of the various fields of the message packet formats are listed in Table 2.16.

Table 2.16 : IGRP Message Format Fields

Field	Description
Version	Sets the value to 0 x 01 always.
Opcode	Sets the value to 0 x 01 for a request and 0 02 for an x update.
Edition	Increments the counter by the sender such that the receiving router knows that it always keeps the latest version of the routing updates.
Autonomous System Number (ASN)	Specifies the IGRP process ID number.
Number of Interior Routes	Indicates the number of routes that are subnets of a directly connected network.
Number of System Routes	Indicates the number of routes for networks from routers within the same AS.
Number of Exterior Routes	Indicates the number of routes from networks outside the ASs.
Checksum	Is calculated from an algorithm based on the header and the entries.
Destination	Indicates the destination network.
Delay	Expresses delay as a multiple of 10 microseconds.
Bandwidth	Defines IGRP bandwidth.
MTU	Expresses MTU in bytes.
Reliability	Specifies error rates of the link.
Load	Specifies load on the link.
Hop Count	Is a number between 0 x 00 and 0 x FF

Enhanced Interior Gateway Routing Protocol

EIGRP is a Cisco proprietary routing protocol. This means that EIGRP does not support routers manufactured by any vendor other than Cisco. The characteristics of EIGRP are:

• Supports VLSM

• Achieves faster convergence than RIP

• Supports multiple routed protocols such as IP, IPX, and AppleTalk

• Requires less bandwidth

• Uses Diffusing Update Algorithm (DUAL)

• Performs load balancing over paths having unequal paths

• Finds alternate routes

EIGRP packets are encapsulated in IP with the protocol field set to 88. The source IP address of the packet is the interface generating this packet. The destination packet can be multicast or unicast depending on the packet type.

The maximum length of the packet is determined by the MTU. The value of MTU varies with respect to the technologies used; it has a default value of 1500 for Ethernet. The different fields of the EIGRP packet are shown in Figure 2.15.

Figure 2.15: The packet format of EIGRP.

The functions of the key fields of the EIGRP packet are:

Version: Indicates the version of the EIGRP process.

Opcode: Specifies the type of the EIGRP packet. Table 2.17 shows the EIGRP packet’s Opcode types.

Table 2.17 : Opcode Types

Opcode	Type
1	Update
3	Query
4	Reply
5	Hello
6	IPX SAP

Checksum: Specifies the checksum of the entire EIGRP packet, excluding the IP header.

Flags: Shows the flag type. An EIGRP packet has two flags. The first bit is called the init bit and is used in a new neighbor relationship. The second bit is the conditional receive bit and is used in a proprietary reliable multicast algorithm.

Sequence Number: Is a 32-bit sequence number. It is used to send messages reliably using RTP.

Acknowledgment Number: Is used to send messages reliably using RTP. It uses a 32-bit sequence number heard from a neighbor to which a packet has been sent.

ASN: Identifies the EIGRP process that is sending the packet. The destination to which the EIGRP packet is being sent will process the packet only if it has an EIGRP routing process with the same number; otherwise, the packet will be rejected.

Type and Length Value (TLV) Field: Follows the EIGRP header. Table 2.18 lists the protocol-specific TLVs.

Table 2.18 : Protocol-specific TLV Types

Number	TLV Types
	General TLV types
0 x 0001	EIGRP parameters (hello/ hold-time)
0 x 0003	Sequence
0 x 0004	Software version
0 x 0005	Next multicast sequence
	TLV Types Specific to IP
0 x 0102	IP internal routes
0 x 0103	IP external routes
	TLV Types Specific to AppleTalk
0 x 0202	AppleTalk internal routes
0 x 0203	AppleTalk external routes
0 x 0204	AppleTalk cable configuration
	TLV Types Specific to IPX
0 x 0302	IPX internal routes
0 x 0303	IPX external routes